Clinical Science and Molecular Medicine (1976) SO, 51-59.

The role of the colon in urea metabolism

in man J. A. GIBSON, N. J. PARK, G. E. SLADEN A N D A. M. DAWSON Department of Gastroenterology, St BarthoIornew's Hospital, London

(Receiued 21 July 1975)

SummarS 1. The urea content of ileostomy effluent has been measured by the urease method as an indirect estimate of the urea concentration in the lumen of the normal ileum. 2. The plasma disappearance of intravenously administered ["Clurea was used to study intestinal urea breakdown. Normal subjects on high and low protein diets and patients with either excised (i.e. with ileostomies) or excluded colons were studied. 3. The 24 h intestinal urea breakdown was considerably greater than the quantity of urea estimated to be entering the colon from the ileum and across the colonic mucosa. 4. Intestinal urea breakdown increased with increase in dietary protein and decreased with, but was not abolished by, exclusion or excision of the colon. 5. Our results suggest that the colonic lumen is not the only site of intestinal ureolysis and that significant quantities of urea must be broken down either at a juxtamucosal site or in the ileum.

ammonia in the intestine daily (Walser & Bodenlas, 1959; Jones, Smallwood, Craigie & Rosenoer, 1969) and the independent finding that an equivalent amount (260 mmol) of ammonia enters the portal circulation from the bowel daily (Summerskill & Wolpert, 1970), it has been suggested that urea is the major substrate for intestinal ammonia production (Summerskill & Wolpert, 1970; Wrong, 1971). Studies of urease activity in man (Aoyagi, Engstrom, Evans & Summerskill, 1966; Evans, Aoyagi & Summerskill, 1966) have shown that, apart from small quantities of mucosal urease in the stomach and small bowel, intestinal urease is bacterial in origin and is located in the colon. We have undertaken the present study to elucidate further the role of the colon in urea hydrolysis and ammonia production. Urea excretion by patients with an ileostomy has been estimated as an indirect measure of the amount of urea entering the colonic lumen and urea turnover has been studied in control subjects and patients with either an excluded or absent colon to elucidate the quantitative role of the colon in urea metabolism.

Key words: colon, urea metabolism. Materials and methods Patients

patients with non-functioning colons. The latter included four Patients who had undergone colonic

Correspondence: Dr J. A. Gibson, Department of Gastroenterology,St Bartholomew's Hospital, London E.C.1.

51

52

J. A . Gibson et al.

exclusion with ileorectal anastomosis for chronic porto-systemic encephalopathy and two patients with ileostomy-one for Crohn’s disease and one for ulcerative colitis. Informed consent was obtained from all subjects. Zleostomy urea

To minimize bacterial ureolysis before estimation, freshly produced ileostomy fluid was immediately cooled by placing it in a container packed in ice and transferred rapidly to the laboratory, where it was diluted with absolute ethanol (three parts fluid to one part ethanol) to inactivate bacteria. The liquid phase was separated by centrifugation through a Centriflow membrane filter (Amicon Corp., Lexington, Mass., U.S.A.) at 200 rev./min for 10 min. Urea concentration was estimated by the urease method. Ammonia concentration before and after incubation with urease was estimated by a modification of the method of Fenton & Williams (1968) with 5 ml of sodium hydroxide (100 mmol/l) for elution and Nessler’s reagent for colour development. The pH of samples incubated with urease was first adjusted to pH 6.6 with Sorensen’s phosphate buffer [ K H z PO4 (0-067 mol/l) adjusted to pH 6.6 with NazHP04 (0-067mol/l)] and 0.1 ml of jack-bean urease suspension (25 mg in 5 ml of ammonia-free water) was added. Incubation was continued for 30 min. The recovery of ammonia after urease incubation was found to be variable, but no systematic error could be found. The possibility that the ethanol used to inhibit bacteria might inhibit urease was considered but specimens diluted with water instead of ethanol showed similar recoveries. Chelating agents were also added in an attempt to remove possible metal inhibitors but these had no effect. Therefore the recovery of ammonia after urease incubation was checked in all specimens by incubating duplicate samples with known added amounts of urea. All results of ileostomy urea were then corrected for this recovery. The reproducibility of this method was assessed by performing serial estimations on two ileostomy specimens, one with a low and one with a high urea concentration. The mean urea concentrations were 0-86 mmolll (SEM 0.23, n = 5) and 4-14 mmol/l (SEM 0.13, n = 8). Finally, bacteriological culture was carried out on a number of the alcoholic mixtures before filtration. These were alwaysfound to be sterile, indicating that spontaneous ureolysis had been effectively stopped.

[ *C]Urea-turnoverstudies

Diet. Control subjects were studied on low (40 g) and high (100 g) protein diets. These diets were isoenergetic at 10-5MJ per day, thechange in dietary protein being balanced by changes in carbohydrate content. Subjects were stabilized on each diet for 7 days before the study, which was carried out on the eighth day. Of the six patients with non-functioning colons, two were studied on the 40 g diet only, with the protocol as above, and one was studied on both diets. Three patients who had recently undergone colonic exclusion for chronic encephalopathy were studied on normal ward diets containing approximately 70 g of protein. Materials. [14C]Urea(The Radiochemical Centre, Amersham, Bucks, U.K.), specific radioactivity 30-50 mCi/mmol, was made up in sterile isotonic sodium chloride (150 mmol/l) at a concentration of 5 pCi/ml. Radioactivity of the standard solution was estimated by adding 10 pl of [14C]urea solution with 0.5 ml Hyamine (500 mmol/l) to 15 ml of scintillation fluid and counting radioactivity for 5 min. 300 r

3

1 50

, 1

1 2

1 3

4

1

1

5 6 Time ( h )

1 7

8

, 9

1 1

, 0

FIG. 1. Semilog regression of 14C concentration in plasma against time showing a straight-line relationship after the 90 min equilibration period.

Experimentalprocedure

Subjects were studied non-fasting, and regular meals were taken throughout the day of study. A steady state was confirmed both by the straight-line regression of log plasma [14C]urea radioactivity against time after at least 90 min had been allowed for equilibration of injected tracer (Walser & Bodenlos, 1959) (Fig. 1) and by the constancy of plasma urea concentration. The dose of [I4C]urea injected for each experiment was approximately 5 pCi, the exact amount being calculated by comparing the weight of the syringe and needle before

,

Role of the colon in urea metabolism and after injection. Immediately before intravenous injection of ['"Clurea the bladder was emptied. Thereafter, timed blood and urine samples were collected at intervals of between 1 and 2 h over a period of 10 h. Total urine volume was measured during each time-peiiod and urine output maintained at at least 2 ml/min by ensuring that subjects drank at least 200 ml of fluidlh. Laboratory procedures

Plasma and urine ["Clurea was estimated by the diacetylmonoxime method on a Technicon 6/60 AutoAnalyzer (Technicon Instruments, New York, U.S.A.). In urine, the high concentration of urea necessitated a tenfold dilution before estimation. To minimize any error introduced by this dilution, both [lZC]ureaand ['"Clurea were estimated in the same sample of diluted urine. Estimation of I4Cwas carried out by a modikation of the method of Walser & Bodenlos (1959). With an Oxford Sampler (Oxford Laboratories International Corp.) paired 1 ml aliquots of either urine or plasma were pipetted into the outer well of 25 ml centre-well flasks. The samples were then acidified with 1 ml of HCl(1 mol/l), a few drops of caprylic alcohol were added, and the flasks shaken to release carbon dioxide. Approximately 1 ml of NaOH (1 mol/l) was added to restore the pH to 7.0 and the solution was buffered to pH 7.2 with 1ml of Sorensen's phosphate buffer [KH2P04(0-067molll) adjusted to pH 7-2 with NazHP04 (0.067 molfl)]. Approximately 1 ml of Hyamine (500 mmol/l) was pipetted into each centre well. Finally, 1 ml of 3% (wlv) urease solution (jack bean urease, Sigma Chemicals, London) was added to the outer well, and the flasks were stoppered with vaccine caps and incubated in a shaking water bath at 37°Cfor 20 min to activate the urease. The outer well was then acidified with approximately 1.5 ml of H1S04 (1 mol/l) injected through the cap. Shaking was continued for a further hour to allow released 14C02to be trapped by the Hyamine in the centre well. Finally, the Hyamine was transferred to counting vials by a multiple washing technique using the scintillationmixture. Five washingsof approximately 1 ml each were found to be adequate. The scintillation mixture used was 15 ml of 2,s-diphenyloxazole (0-3 %) and 1,4-bis-(5-phenyloxazol-2-yl)-benzene (0.01%). Radioactivity of samples was counted for

53

20 min and at least 1500 counts using a Tracer-Lab Corulmatic 200 Ambient Temperature liquidscintillation counter. Results were recorded through a linked Diehl computer, programmed to correct for quenching, and results expressed in d.p.m. Calculation Calculation of urea dynamics was initially carried out by the method of Walser & Bodenlos (1959). It should be noted that in their studies urinary delay time varied from 1-6 h to zero. It is clear therefore that any error resulting in an underestimation of delay time could result in a negative value making further calculation invalid. Such values were obtained in some of om studies. A review of methodology showed that recovery of standard solution of ['"Clurea added to control plasma or diluted urine, and extracted as described, was over 99%. It is thought that any error was probably the result of incomplete bladder emptying or possibly in estimation of [lZC]urea in urine because of the high urinary urea concentration already mentioned. We have therefore used an alternative method of calculation (Deane, Desir & Umeda, 1968), which gives results comparable with those of other workers while avoiding the calculation of urinary delay time. In this method the urea volume of distribution ( V ) is calculated at timed intervals after the injection of ['"Clurea from the formula: Injected dose of ['"Clurea -cumulative - urinary ['"CJurea excretionup to time t - ['"C]urea concentration in plasma at time t Plasma [14C]urea concentration decreases faster than [14C]urea is being excreted in the urine, indicating that urea is being lost by a second route, that is breakdown in the gut. Thus the apparent volume of distribution will increase with time. The reciprocal of V,,,,. plotted semilogarithmically against time gives a straight-line relationship with a negative slope if time is allowed for complete distribution of injected ['*C]urea in the native urea pool (Fig. 2). The slope of this line indicates the fraction of the residual [14C]ureabroken down in the gut each hour, which, after equilibration, will be equal to the fraction of the non-labelled urea pool broken down in the gut each hour. As stated, this method makes no specifx allowance for the factoks of excess excretion or delay of urinary

V,,,ea at time t

54

J. A . Gibson et al.

excretion. However, from the graph in Fig. 2 it can be seen that, at any time t : 1 [14C]ureaconcentration in plasma at time r e V,,,., Injected dose of [l4C1urea - cumulative urinary excretion of [l4C1ureaup to time t

Excess excretion will be contained in the cumulative urinary excretion of I4C up to time t. Plasma concentrations of [14C]urea are only considered after equilibration, by which time excess excretion will have taken place. Excess of excretion will iniluence the intercept of the curve but the slope

were, however, relatively small (between 1% and 2%) and could be ignored. From this method the urea pool is the product of V,,., and plasma urea concentration, and the breakdown of urea in the gut is calculated from the product of the urea pool and the slope of the line in Fig. 2. Urinary urea excretion is calculated from the excretion of [“Clurea in the urine. In the steady state it is assumed that: Urea production = urea excretion in urine+urea breakdown in the gut Statistical comparison of mean values in studies within and between individuals was carried out with Student’s t-test.

Results Ileostomy urea

Ileostomy urea concentration in the twenty-nine specimens varied between 0.58 and 9.17 mmol/l. In twenty-six studies the blood urea was estimated at the time of ileostomy sampling and the ratio of

(b)

I

.

0

1

2

3

4

5 6 Time ( h 1

7

8

9

1

0

Fro. 2. Semilog regression of 105/Vur., against time in one control subject (J.A.G.) on the low (a: 40 g/day) and high (b: 100 g/day) protein diets.

will be unaffected. Thus the fractional rate of urea breakdown will be valid, but because the urea space will be overestimated the absolute rate of breakdown may be rather too large. Urinary delay time implies that cumulative urinary excretion of I4C will underestimate the total amount of 14C which has left the circulation by time r. This could affect both slope and intercept and so introduce errors into the calculation of urea pool and fractional metabolism. The importance of this can be assessed by replotting the data, assuming a delay time of 1-0 h (the mean value derived by Walser t Bodenlos, 1959). In two studies replotting the curves increased the intercept by a mean of 7%, i.e. a 7% underestimation of the volume of distribution and hence urea pool. The slope was increased 8.5% on average. The resulting changes in urea production and gut urea breakdown

-

0

FIG. 3. Ratio of urea concentration in ileostomy fluid and blood. Blood was taken at the time of each ileostomy fluid

sampling and the ratio of ileostomy urea to blood urea estimated.

Role of the colon in urea metabolism

8 8

B

55

J. A. Gibson et al.

56

ileostomy urea to blood urea calculated. This varied between 0-13 and 1.90, mean 0.715 (SD 0.40, n = 26) (Fig. 3). Urea metabolism

Normal subjects. Results of studies in six normal subjects are shown in Table 1. The mean breakdown of urea in the gut increased from 4.3 mmol/h (SEM 0.49, n = 6) on a low protein diet to 7 6 mmol/h (SEM 0.89, n = 6) on a high protein diet. Paired testing showed that this increase is statistically significant (t 3,938, PP>0.6). That is to say, in the subjects studied with normal renal function, a constant proportion of synthesized urea is broken down in the gut. Patients without functional colons. The results of urea-turnover studies in subjects with non-functioning colons is shown in Table 2. The variable dietary protein intake in these patients makes it impossible to compare gut urea breakdown in absolute amounts with breakdown in normal control subjects. However, we have already shown in the control group that gut urea breakdown is a constant percentage of

urea production unaffected by dietary protein. If this percentage is compared in the two groups it will be seen that removal of the colon results in a fall from a mean value of 28.3% (SEM 1.87, n = 6) in control subjects to 19.7% (SEM 2.6, n = 6) in patients with non-functioning colons. By unpaired t-testing this difference is significant at the 2 5 % level (t 2693, P